Staphylococcus aureus (SA) is the most common cause of skin and soft tissue infection. Despite multiple classes of antibiotics with in vitro actiity against even the methicillin resistant S. aureus (MRSA) strains, SA infections are often persistent and reactivate after long periods of latency. As therapeutic strategies targeting the organism are not successful in eradicating these infections, the goal of this project is to target the host - to identify components of the keratinocyte that are exploited and serve as a nidus for persistent SA infection. We postulate that the autophagosome of keratinocytes represents such a protected niche for the selection of SA mutants that have adapted to the conditions of oxidative stress within these cells and tolerate local concentrations of toxic polyamines. In the R21 phase of this project, we will fully characterize the intracellular compartment associated with SA persistence and identify the signaling pathways activated by SA to induce autophagosome formation. As autophagy normally functions to eradicate intracellular pathogens, we predict that specific SA mutants are selected in response to this milieu that avoid lysosomal fusion and eradication and can tolerate the oxidative stress within this compartment. We will use a bioinformatics approach, comparing whole genome sequencing data from laboratory strains selected for adaptation to the autophagosome with the genotypes of clinical isolates from chronic infections. The goals of the R21 component are to establish that specific SA mutants are actively selected in response to the autophagosome and establish a nidus of latent infection within keratinocytes. In the R33 component, we will develop a model of human skin infection; first using organotypic cultures of human primary keratinocytes, to establish that pharmacologic and genetic techniques to impair the induction of autophagy in keratinocytes will prevent the establishment of chronic SA infection. Using drugs that target the AMPK-TORC1 pathway or siRNA inhibiting specific autophagy genes, we expect to block autophagy and prevent the selection of SA mutants that can survive in this setting. A SCID:hu mouse with human skin xenografts will then be used as proof of principle in the setting of human skin infection in the presence of an intact immune response. The most effective reagents identified in vitro will be tested on human skin grafts for their ability to prevent the adaptation of SA to the human keratinocyte.